Table 8. mycoSORT Results - Set of Features F1+F2+F3+F4.
Under-sampling(USF) | Classifier | Precision | Recall | F-measure | MCC | F-2 |
Training set with USF 0% | Naive Bayes | 0.355 | 0.727 | 0.477 | 0.431 | 0.600 |
Training set with USF 0% | LMT | 0.685 | 0.420 | 0.521 | 0.498 | 0.460 |
Training set with USF 0% | LibSVM | 0.867 | 0.087 | 0.158 | 0.257 | 0.110 |
Training set with USF 5% | Naive Bayes | 0.365 | 0.740 | 0.489 | 0.446 | 0.610 |
Training set with USF 5% | LMT | 0.585 | 0.480 | 0.527 | 0.484 | 0.500 |
Training set with USF 5% | LibSVM | 0.729 | 0.287 | 0.411 | 0.424 | 0.330 |
Training set with USF 10% | Naive Bayes | 0.349 | 0.787 | 0.484 | 0.448 | 0.630 |
Training set with USF 10% | LMT | 0.552 | 0.600 | 0.575 | 0.526 | 0.590 |
Training set with USF 10% | LibSVM | 0.670 | 0.420 | 0.516 | 0.491 | 0.450 |
Training set with USF 15% | Naive Bayes | 0.342 | 0.787 | 0.477 | 0.441 | 0.620 |
Training set with USF 15% | LMT | 0.478 | 0.647 | 0.550 | 0.498 | 0.600 |
Training set with USF 15% | LibSVM | 0.607 | 0.473 | 0.532 | 0.491 | 0.490 |
Training set with USF 20% | Naive Bayes | 0.342 | 0.793 | 0.478 | 0.443 | 0.630 |
Training set with USF 20% | LMT | 0.425 | 0.64 | 0.511 | 0.456 | 0.580 |
Training set with USF 20% | LibSVM | 0.521 | 0.587 | 0.552 | 0.500 | 0.570 |
Training set with USF 25% | Naive Bayes | 0.322 | 0.787 | 0.457 | 0.421 | 0.610 |
Training set with USF 25% | LMT | 0.389 | 0.747 | 0.511 | 0.469 | 0.630 |
Training set with USF 25% | LibSVM | 0.474 | 0.667 | 0.554 | 0.504 | 0.620 |
Training set with USF 30% | Naive Bayes | 0.336 | 0.773 | 0.469 | 0.430 | 0.610 |
Training set with USF 30% | LMT | 0.398 | 0.780 | 0.527 | 0.490 | 0.650 |
Training set with USF 30% | LibSVM | 0.459 | 0.673 | 0.546 | 0.496 | 0.620 |
Training set with USF 35% | Naive Bayes | 0.304 | 0.800 | 0.440 | 0.406 | 0.600 |
Training set with USF 35% | LMT | 0.343 | 0.760 | 0.473 | 0.433 | 0.610 |
Training set with USF 35% | LibSVM | 0.357 | 0.793 | 0.493 | 0.458 | 0.640 |
Training set with USF 40% | Naive Bayes | 0.295 | 0.780 | 0.428 | 0.389 | 0.590 |
Training set with USF 40% | LMT | 0.361 | 0.847 | 0.506 | 0.481 | 0.670 |
Training set with USF 40% | LibSVM | 0.331 | 0.793 | 0.468 | 0.433 | 0.620 |
Results of Positive Class on Feature Setting #4, Using Bio-entities, Content and EC Numbers as Features.